spark-recipe-delegate-and-track
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data by reading email threads, which creates a surface for indirect prompt injection attacks. \n- Ingestion points: The
spark thread <id>command in Step 3 ingests email bodies into the agent context. \n- Boundary markers: There are no instructions to use delimiters or ignore instructions found within the email content. \n- Capability inventory: The skill utilizesspark actioncommands to assign emails, set due dates, and change delegation status. \n- Sanitization: The instructions do not specify any sanitization or validation of the email content before it is processed by the agent.
Audit Metadata